Time Of Observation Bias

Time of observation bias affects thermometer records. In this short video I show how to use visitech.ai to visualize and understand a difficult scientific problem.

About Tony Heller

Just having fun
This entry was posted in Uncategorized. Bookmark the permalink.

6 Responses to Time Of Observation Bias

  1. Bob G says:

    congratulations Tony for exposing another scam pulled on us by the climatenistas

  2. conrad ziefle says:

    I always wondered just how much time of observation did bias readings, and just how well documented the actual time of observation is. Also, if the time of observation is consistently the same, then how important is it, since all of the data has the same reference point.

    • conrad ziefle says:

      I didn’t think it would make that much difference to the average, but it does! You would think that they would be smart enough to know that they are causing a double count error, and probably they did, but, in those days, no one really thought about climate change being a big hocus-pocus. And two, the guy taking the reading left at 5pm and wasn’t about to come back later to reset the thermometer. If they reset it at 7-8 am when they came in, then it would have worked out better, but they probably were dealing with other important matters at 7am, like gathering around the water cooler.

      • conrad ziefle says:

        So how is NOAA adjusting the data? Clearly, some places have a double count errors and others don’t. Do they go through station by station to find the errors, or do they blanket adjust them all downward?

  3. Crashex says:

    The example you presented leads to a double count of a max temp across two days, resulting in an improperly high average temp. So, the correction would be do reduce the average temperature by some statistical factor to account for the error.

    But, the temperature adjustment chart you presented first INCREASES the average temperature?? And a similar time of observation issue could be expected for morning observations regarding the daily mins.

    All of it is part of the min/max observation records and describing the resulting median as the average–not the same thing.

    • conrad ziefle says:

      I’m not sure if it is true everywhere or just here, but the minimum temperature here is roughly around sunrise or slightly before, but since with the old-fashioned thermometers you are picking two temperatures, one off of each end, an intelligent person could set a schedule, maybe not foolproof, but which would work 99% of the time for thermometer readers in his neck of the woods. The issue, I think, is that this is no longer a problem with modern equipment because you can take the exact temperature every millisecond, if you want, and then a computer can pick the ones you want, average, medium, or any statistic you want. The problem is how do you fix old data that was taken by hand. I think there are ways that make sense. Using the neighboring station which clearly did not have a double peak is one way, and probably with AI or a sophisticated program you can probably identify where these have occurred, and fix only those specific ones.

Leave a Reply to Crashex Cancel reply

Your email address will not be published. Required fields are marked *